Testing the fidelity of two paleoproductivity proxies in the context of

Top down and bottom up:
Testing the fidelity of two paleoproductivity proxies in the context
of the late Miocene to early Pliocene “biogenic bloom”
PI: Katharina Billups, University of Delaware
Funded by: American Chemical Society – Petroleum Research Fund, 2005
Top down and bottom up: Testing the fidelity of two paleoproductivity proxies in the context of
the late Miocene to early Pliocene “biogenic bloom”
Abstract
I propose to compare two paleoproductivity indicators recording “opposite” aspects of
primary productivity, organic carbon production at the sea surface (coccolith Sr/Ca ratios) and its
consumption on the sea floor (benthic foraminiferal accumulation rates, BFAR). Calcification and
growth rates of coccolithophores can be linked to particulate organic matter production in the photic
zone (Balch and Kilpatrick, 1996). Benthic foraminifera feed on particulate organic matter and thus
respond to the flux of organic carbon reaching the deep ocean (Berger and Wefer, 1990). If coccolith
calcification rates are indicative of surface ocean productivity as recorded in coccolith Sr/Ca, and a
significant amount of the particulate organic matter makes it to the sea floor as recorded by BFAR,
then a relationship should exist between coccolith Sr/Ca ratios and benthic foraminiferal
accumulation rates. I will construct down-core coccolith Sr/Ca records to parallel down-core BFAR
during the late Miocene to early Pliocene “biogenic bloom” interval. I have chosen this interval
because there is benthic foraminiferal evidence for an increase in paleoproductivity; hence there is a
signal. Each proxy has its own set of biological and biogeochemical controls, and consequently
uncertainties. This study will test the applicability of the two proxies to paleoproductivity
reconstructions on longer time scales in the face of these uncertainties. Results of this study are
important to the Petroleum Research Fund because an improved understanding of these
paleoproductivity proxies will be valuable to the reconstructions of primary marine productivity, the
ultimate determinant of potential hydrocarbon resources.
1
Goal
To explore the coherence of two paleoproductivity indicators: Coccolith Sr/Ca ratios as a
measure of algal growth and calcification rates at the sea surface and benthic foraminiferal
accumulation rates as a measure of organic carbon consumption on the sea floor (Figure 1)
Figure 1. Cartoon to illustrate an oversimplified relationship between organic matter production (exemplified here
by coccolithophores) and consumption at the sea surface (e.g., zooplankton respiration), continued consumption and
transfer through the water column, and consumption on the ocean floor (e.g., by benthic foraminifera).
2
1. Introduction
1.1 Top Down:
Stoll’s and Rickaby’s research has shown that coccolith Sr/Ca ratios are in part a function of
coccolithophorid calcification and growth rates. Coccolithophorid culture experiments demonstrate
that the distribution coefficient of Sr (DSr) in coccolithophorid calcite is positively correlated with
rates of organic carbon fixation and calcification (Stoll et al., 2000a; 200b; Rickaby and Schrag,
2002). Calcification and growth rates of coccolithophores can be linked to particulate organic matter
production (Balch and Kilpatrick, 1996), thus coccolith Sr/Ca ratios should provide a proxy for
primary productivity. This is supported by spatial variations of core-top coccolith Sr/Ca ratios in
concert with primary productivity across the eastern equatorial Pacific (Stoll and Schrag, 2000)
(Figure 2). In down-core studies, surface water productivity has been implicated to explain temporal
variability in Cretaceous bulk sediment (primarily coccoliths) Sr/Ca ratios (Stoll and Schrag, 2001)
and Paleocene/Eocene coccolith Sr/Ca data (Stoll and Bains, 2003). Over Cenozoic time scales, it is
possible to relate common long-term trends in bulk sediment Sr/Ca ratios from various regions with
what is known about large-scale productivity changes (Billups et al. 2004).
110W
2.35
productivity
2.30
coccolith Sr/Ca
900
2.25
800
2.20
700
2.15
2.10
600
2.05
500
Coccolith Sr/Ca (mmol/mol)
Productivity (mgCm-2day-1)
1000
2.00
-10
-5
0
5
10
Latitude (N)
Figure 2. Relationship between primary productivity and coccolith Sr/Ca ratios in the eastern equatorial Pacific
(data from Stoll and Schrag, 2000). Coccolith Sr/Ca ratios increase with surface ocean productivity, which is
consistent with the positive correlation between the distribution coefficient of Sr in coccolith calcite and
calcification rates (Stoll et al., 2002a; b; c; Rickaby and Schrag, 2002).
3
Although existing evidence suggests that coccolith Sr/Ca ratios provide a geologic
perspective on the primary production of organic matter, there are uncertainties when applying this
method to longer time scales. These include secular variations in seawater Sr/Ca (Lear et al., 2003),
diagenesis (e.g., Hampt-Andreasen and Delaney, 2000), temperature effects on the coccolith Sr/Ca
ratios (e.g., Stoll et al., 2000a), and species-specific Sr/Ca ratios (Stoll and Bains, 2003). Further
concerns may arise from changes in surface and deep water carbon chemistry (pH effects on
calcification rates and dissolution, respectively) as these are likely to change in response to primary
productivity.
It is my objective in this study to constrain as many of these uncertainties as possible. As
detailed in the Research Strategy seawater Sr/Ca variations can be accounted for by calculating the
DSr from the coccolith Sr/Ca ratios and the seawater Sr/Ca curve (at least to the extent of which
seawater Sr/Ca ratios are accurate). SEM imaging can be used to look for obvious signs of diagenetic
overprinting of the individual coccoliths. Surface water temperatures can be constrained using
planktonic foraminiferal Mg/Ca ratios. Species-specific Sr/Ca ratios can be limited by separating the
samples into discrete size fractions. Down-core nannofossil assemblages can be monitored to obtain
some measure of the extent of species-specific bias.
The geochemical concerns are more difficult to constrain. To date, there is no evidence for a
relationship between coccolith Sr/Ca ratios and culture water pH (Stoll et al., 2002b). Thus, although
productivity induced changes in pH may affect calcification rates, this mechanism may not drive
coccolith Sr/Ca ratios. Regarding dissolution, unlike foraminiferal calcite, coccolith Sr/Ca ratios do not
appear to be sensitive to partial dissolution because Sr is homogeneously distributed throughout the
calcite platelet (Stoll et al., 2002b). In light of these biogeochemical and physiological controls, it is
important to establish whether down-core changes in coccolith Sr/Ca ratios are dominated by a
productivity signal.
4
1.2 Bottom up:
Benthic foraminifera feed on particulate organic matter settling from the photic zone and should
therefore respond to export production, or rather, the flux of organic carbon reaching the deep ocean.
Numerous studies from different regions have illustrated that there is an increase in benthic
foraminiferal numbers with increased carbon flux to the sea floor (e.g., Berger and Wefer, 1990;
Herguera and Berger, 1991; Nees, 1997; Schmiedl and Mackensen, 1997; Yasuda, 1997; van der
Zwaan et al., 1999; Herguera, 2000; Diester-Haas et al., 2004). For a comprehensive review see the
special issue of Marine Micropaleontology (2000) on Foraminiferal Proxies and Paleoproductivity).
In this issue, Herguera (2000) illustrates that a well-defined relationship exists between benthic
foraminiferal accumulation rates (BFAR) from core top sediments from the equatorial Pacific and
Organic Carbon flux to the sea floor (mgC/cm2kyr)
Atlantic and the corresponding organic carbon flux (Figure 3).
250
Western eq. Pacific
Eastern eq. Pacific
Atlantic transect
200
150
100
50
Organic carbon flux to the sea floor = 6.5BFAR6.4
0
0
50
100
150
200
250
2
BFAR (no/cm kyr)
Figure 3. Relationship between core top benthic foraminiferal accumulation rates and organic carbon flux to the sea
floor from Herguera (2000). The organic carbon flux numbers are derived from the regression between organic
carbon flux, water depth and primary productivity of Berger and Wefer (1990). Herguera (2000) show how the
BFAR can then be used to reconstruct paleo primary productivity (pPP) via the observed relationship between
BFAR and water depth (Z): pPP = 0.4*Z*BFAR0.5. (eq. 1).
Berger and Wefer (1990) have established that there is a relationship between primary
productivity at the sea surface and the organic carbon flux to the sea floor. Using this regression (not
5
shown here), Herguera (2000) then derives a simple equation that relates “paleo” primary
productivity to BFAR (equation 1):
pPP = 0.4*Z*BFAR0.5 (eq. 1)
where pPP stands for paleo primary productivity, Z is the water depth of the sample, and BFAR are
the benthic foraminiferal accumulation rates.
Modern ocean studies, however, indicate that the transfer of organic matter from the surface
to the deep ocean is complex. The amount of export production of organic carbon out of the photic
zone is not simply a function of the magnitude of primary productivity at the sea surface (Lampitt
and Antia, 1997). Phytoplankton community structure is an important control on the amount of
carbon leaving the photic zone (Boyd and Newton, 1999). For example, Francois et al. (2002) argue
that carbonate producers provide a more efficient means to transfer carbon into the deep ocean owing
to a “ballasting effect” of relatively heavy carbonate. Thus carbon in carbonate producing regions is
transferred to deeper depths before it is remineralized than carbon in biogenic opal producing
regions. These authors point out that while less organic carbon may be produced in the photic zone of
an oligotrophic carbonate producing region than in a nutrient rich biogenic silica dominated region, a
relatively larger fraction of it can make it to the sediment/water interface. Application of BFAR to
reconstructions of paleoproductivity, which are calibrated to a few specific sites representing a
specific environment, must assume that these factors remain constant through time.
Perhaps the largest source of uncertainty applying BFAR to paleoproductivity reconstruction
arises from uncertainties associated with age control. Benthic foraminiferal accumulation rates are
derived from mass accumulation rates and hence depend on sedimentation rates. Sedimentation rates
are usually assumed linear between control points, which can be separated by as much as millions of
years if based on biostratigraphy, or as little as a few thousand years if derived from astrochronology.
However, age models are subject to change, which introduces unpredictable uncertainties in proxies
dependent on sedimentation rates.
6
1.3 Top down and bottom up:
Coccolith Sr/Ca and benthic foraminifera offer two “opposite” views of primary productivity,
production at the sea surface and consumption on the ocean floor. Each proxy has its own set of
biological, biogeochemical, or geological controls, and hence uncertainties. Coupling coccolith Sr/Ca
ratios with BFAR should present a coherent picture of paleo primary productivity if 1) coccolith
Sr/Ca ratios are tightly coupled to changes in primary productivity regardless of the hydrographic
setting, and 2) BFAR-based reconstruction of the deep ocean organic carbon flux faithfully tracks
temporal changes in surface ocean productivity. In this project I propose to construct down-core
records of coccolith Sr/Ca ratios from two sites for which corresponding BFAR data are available to
test the extent to which it is possible to interpret these two different proxies in a common
biogeochemical framework, primary productivity. A lack of a coherent picture would highlight the
complex nature of productivity, carbon flux in the deep ocean, and the reconstruction thereof.
Consistent trends, however, would strengthen the application of these proxies to paleoproductivity
reconstructions. The results of this study will test our understanding of these two proxies.
2. Late Miocene/early Pliocene “Biogenic Bloom”
The late Miocene/early Pliocene “biogenic bloom“ interval (~7-3 Ma) provides an opportune
time interval to examine the temporal relationship between coccolith Sr/Ca ratios and BFAR because
a large body of evidence exists showing that during this interval of time an increase in primary
productivity occurred at a number of sites in the Pacific and Indian Oceans (Petersen et al., 1992;
Berger et al., 1993; Farrell et al., 1995; Dickens and Owen, 1994; Grant and Dickens, 2002) and the
Atlantic (Diester-Haass et al., 2002; 2004; Diester-Haass et al., submitted). Thus a great opportunity
now exists to test our understanding of the coccolith Sr/Ca and benthic fauna productivity proxies in
the context of late Miocene through early Pliocene climate change. Causes of the "biogenic bloom"
7
are being debated, but these are not the focus of this study (hence the lack of a more detailed review
of the evidence here). What is important here is that a mechanism exists that, according to the
conventional interpretation of the proxies, should change both coccolith Sr/Ca ratios and BFAR.
3. Detailed Objectives
• Construct size fraction specific records of coccolith Sr/Ca ratios from ODP Site 982 (North
Atlantic) and from ODP Site 925 (western tropical Atlantic) using the same intervals as used
for benthic foraminiferal accumulation rates (from Diester-Haass et al., submitted);
• Calculate the DSr from the Sr/Ca ratios and published seawater Sr/Ca ratios to account for secular
changes in seawater Sr and Ca variations;
• Construct parallel records of planktonic foraminiferal Mg/Ca to monitor surface water
temperature changes;
• Establish down-core patterns of nannofossil species assemblages at both sites;
• Compare the coccolith Sr/Ca records to the benthic foraminiferal accumulation rate/
paleoproductivity records from the same intervals;
• Evaluate if, to what degree, or which coccolith size fraction best parallels changes in benthic
foraminiferal derived paleoproductivity at each site.
4. Research Strategy
My strategy falls into two catogories. In 4.1 Site Selection, I justify the choice of the two
sites. These include the availability of benthic foraminiferal data, sedimentological criteria,
hydrographic regime and age control. In 4.2 Coccolith Sr/Ca Ratios, I specify how I address caveats
associated with the geochemical proxy, namely secular changes in seawater Sr/Ca ratios, temperature
effects and changing coccolith species assemblages.
8
4.1 Site Selection
Paleoproductivity signal
I have chosen ODP Sites 982 (subpolar North Atlantic) and ODP Site 925 (western tropical
Atlantic) (Figure 4) because counting of benthic foraminifera is complete indicating that an increase
occured in paleoproductivity at both sites, confirming that a signal exists (see Preliminary Results,
page 15). This provides the basis for the next step, and I have received the bulk sediments from each
of the intervals to construct parallel records of coccolith Sr/Ca ratios.
982
*
925
*
Figure 4. Location of ODP Site 925 (western tropical Atlantic, 3053 m water depth) and Site 982 (North Atlantic,
1134 m water depth). The shading reflects the annual average phosphate content, which I chose purely for
illustrative purposes, ranging from 0.2 μmolar (dark gray) 0.4 μmolar (light gray). The map was generated using the
Lamont Doherty data library.
Carbonate content and diagenesis
ODP Sites 982 and 925 have a high carbonate content (averaging ~90% and 80%,
respectively Shipboard Scientific Party, 1996; 1995). This is important because Stoll and Schrag
(2000) demonstrate that there can be an effect of adsorbed Sr and Ca cations from alumino silicates
in the coccolith size fraction on the Sr/Ca ratio of the sample. By mass balance this effect decreases
with increasing carbonate content (Stoll and Schrag, 2000). Thus high carbonate content (>80%) is
critical for unambiguous results.
9
High carbonate content, however, is an environment conducive to carbonate diagenesis (e.g.,
Stout, 1985; Hampt-Andreasen and Delaney, 2000). It may not be possible to entirely rule out such
effects, but SEM imaging of coccoliths from both sites does not show any obvious sign of diagenetic
overprinting (Figure 5). Neither of the imaged samples shows evidence for euhedral calcite crystals
that would indicate significant recrystallization.
Site 982
Site 925
Figure 5. Mid core examples of coccoliths from North Atlantic Site 982 (Calcidiscus) and western tropical Atlantic
Site 925 (Umbilicosphaera). The calcite platelets are well preserved; there is no evidence of significant secondary
calcification.
Hydrography
The two sites represent different hydrographic regimes (Figure 4), but only Site 925 lies
within the BFAR-paleoproductivity calibration region of Herguera (2002). In the modern ocean large
seasonal differences in primary productivity exist at subpolar Site 982 (0.1-0.65 gC/m2/day), and at
tropical Atlantic Site 925 primary productivity varies little throughout the year (0.2-0.35 gC/m2/day)
although the annual averages are similar (12.5 gC/m2/day; Antoine et al., 1996). I do not propose to
use the modern setting as an analog for the late Miocene/early Pliocene, but I can assume that the
hydrographic regime differed from each other in the past simply because of their high versus low
latitude locations. Thus, choice of these two sites tests the applicability of the BFAR based
productivity proxy outside the environment for which it is calibrated.
10
Sites 982 and 925 come from different water depths (3053 m vs. ~1134m, respectively),
which can be accounted for by applying the regression of Herguera (2000) and calculating
paleproductivity from BFAR (equation 1, page 5). However, the approach does assume that the water
depth has remained constant through time. At both sites, the paleo water depth during the middle to
late Miocene may have been shallower by 100-150 m than the modern water depth (Site 982:
Andersson and Jansen, 2003; site 925: Shipboard Scientific Party, 1995). The water depth effect will
introduce an uncertainty of about 4 % in the paleoproductivity estimates, which is not large enough
to mask any longer-term trends.
Age control
Excellent age control exists at both sites, which is important for the calculation of sediment
accumulation rates and hence BFAR from the number of benthic forams per gram sediment. At Site
982 the age model is based on tuning a benthic foraminiferal δ18O record to obliquity (Hodell et al.,
2001). Thus Site 982 age control points are denser (40 kyr) than the temporal resolution in this study
(~50-80 kyr), and the assumption of constant sedimentation rates between control points is most
likely a good one. At Site 925, the time scale is based on magnetic susceptibility tuned to Northern
Hemisphere summer insolation (Shackleton and Crowhurst, 1997; Shackleton and Crowhurst, pers.
communication). Again, age control points are denser than the sample spacing of this study and
sedimentation accumulation rates can be assumed linear between control points.
4.2 Coccolith Sr/Ca Ratios
Seawater Sr/Ca ratios
Secular changes in seawater Sr/Ca ratios introduce a trend into the geochemical time series.
There is evidence from benthic foraminiferal Sr/Ca ratios that seawater Sr/Ca ratios changed during
11
the late Miocene to early Pliocene (Lear et al., 2003) (Figure 6). The long-term decrease toward 7 Ma
and subsequent increase toward the present needs to be removed from the coccolith Sr/Ca time
series. This will be achieved by calculating the distribution coefficient of Sr in coccoliths (equation
2) based on the measured coccolith Sr/Ca ratios and the benthic foram derived seawater Sr/Ca ratios
of Lear et al. (2003):
DSr = [Sr/Cacoccolith]*[Sr/Caseawater]-1
(equation 2)
Thus I will be comparing temporal changes in the DSr with benthic foraminiferal accumulation
rates/paleoproductivity, not simply the coccolith Sr/Ca ratios.
10.0
Sr/Caseawater (mmol/mol)
9.5
9.0
8.5
8.0
7.5
7.0
0
1
2
3
4
5
6
7
8
9
10
Age (Ma)
Figure 6. Seawater Sr/Ca ratios calculated from benthic foraminiferal Sr/Ca ratios using a benthic foraminiferal
distribution coefficient of Sr of 0.165 (Lear et al., 2003). Benthic forams may be the best indicator of seawater ratios
because the only known variable to affect the DSr is water depth, which has been accounted for in this study.
Seawater Sr/Ca ratios change during the time interval under investigation (shaded) highlighting the importance of
eliminating the long-term trend in the coccolith Sr/Ca records.
Temperature effects:
There is a temperature effect on Sr/Ca ratios of about 2 % ºC-1 (Stoll et al., 2002b). In a
tropical setting, such as Site 925, where even uppermost sea surface temperatures are comparatively
stable (e.g., the modern seasonal range at 0 m water depth is only about 2ºC, Levitus and Boyer,
12
1994) this issue may be a minor concern (assuming that the hydrographic regime of the past is
similar to today). At subpolar North Atlantic Site 982, on the other hand, the modern seasonal sea
surface temperatures change is large, about 6ºC, which alone would introduce a ~12% change in
Sr/Ca ratios. However, being photosynthetic, coccolithophores occupy the entire photic zone, thus
the modern seasonal sea surface temperatures may be an overestimate of the photic zone temperature
range. For example, at only 50 m water depth the seasonal temperature changes is strongly attenuated
(~2ºC Levitus and Boyer, 1994).
Because it is difficult to argue that seasonal temperature changes are a quantitative analog for
down-core variations, I will use planktonic foraminiferal Mg/Ca ratios to constrain the possibility of
sea surface (or upper ocean) temperature changes that have affected the coccoliths Sr/Ca ratios. The
temperature dependency of Mg incorporation into planktonic foraminiferal calcite has been well
established and by now a number of species specific calibrations exist (e.g., Dekens et al., 2003;
Anand et al., 2003). At western tropical Atlantic Site 925 I will use the abundant Globigerinoides
sacculifer, a foraminifera that primarily occupies the mixed layer. At North Atlantic Site 982, I will
use Globigerina bulloides, a mixed layer and thermocline dweller. The reconstructed temperature
variability may overestimate the temperatures of the coccolithophorid calcification environment,
particularly if the photic zone is deeper than the thermocline, but they will provide an important
upper constraint.
In this context, secular changes in seawater Mg/Ca ratios need to be considered. However,
although seawater Mg/Ca ratios may have been lower than today during the late Miocene/early
Pliocene, no rapid changes occurred during this time interval (Wilkinson and Algeo, 1989; Stanley
and Hardie, 1998). Thus, while the absolute temperature values may be affected by different seawater
ratios, the variability in foraminiferal Mg/Ca ratios between sampling intervals, ~80 kyr, is due to
temperature (the residence time of Mg and Ca being 13 Myr and 1 Myr, respectively). I believe that
this aspect of the project is ideally suited for an undergraduate senior thesis.
13
Coccolith species effects
Stoll’s recent work shows that there may be species-specific offsets in coccolith Sr/Ca ratios.
Stoll and Bains (2003) illustrate that in Paleocene sediments, coccoliths belonging to larger size
fractions (8-12 μm) have lower Sr/Ca ratios than smaller taxa (3-5 μm). Size division may split the
nanno fossils broadly based on ecology with most of the cooler water/mesotrophic species in the
small range, and the warm water/oligotrophic species in the larger size range (e.g., Gibbs et al.,
2004). Thus temporal changes in species assemblages, perhaps in response to surface water
paleoproductivity, will affect the interpretation of the Sr/Ca records.
We will separate the coccolith fraction into two size fractions by microfiltering at 8 μm.
Although we cannot constrain species assemblage changes within each size faction, we will be able
to eliminate bias due to changes in small versus large species, which may be brought about by
environmental changes, throughout the study interval. Hence, Sr/Ca ratios within each size fraction
more closely reflect the geochemical signal due to the kinetics of Sr incorporation.
In addition, we will prepare smear slides and record changes in coccolith species abundance
in each of the intervals to be measured. Samantha Gibbs, currently at nearby Penn State, is an expert
on nannofossil taxonomy and assemblages and has offered her expertise (see attached letter of
support). Graduate student Amanda Waite, who is beginning this project for a Master’s Thesis, has
already visited Samantha Gibbs to learn coccolith identifications. The preliminary look tells us that
there are comparatively few nannofossil species at North Atlantic Site 982 and that abundances are
relatively constant. At tropical Atlantic Site 925, however, species diversity is high, and as
demonstrated by the recent early Pliocene study of Gibbs et al (2004), assemblages change on orbital
time scales. Thus, monitoring the down-core species assemblages will be particularly important at
Site 925.
14
5. Preliminary Results
As noted above, BFAR have already been determined at both sites (Figure 7A). The results
indicate that Site 982 BFAR are higher than those at Site 925, which is an example of the water depth
effect (e.g., Herguera, 2000). Converting BFAR to paleoproductivity using equation 1 indicates more
similar data averages (Figure 7B). The amplitude of the two major productivity peaks is comparable,
6000
North Atlantic Site 982
BFAR (no/cm2kyr)
A
Tropical Atlantic Site 925
5000
50
B
40
4000
30
3000
20
2000
modern
1000
10
0
Paleoproductivity (gC/cm2kyr)
although the timing differs (at 5 Ma and 6.5 Ma at sites 982 and 925, respectively).
0
3
4
5
6
7
8
9 3
Age (Ma)
4
5
6
7
8
9
Age (Ma)
Figure 7. Benthic foraminiferal accumulation rates (BFAR) (A) and calculated paleoproductivity (B) from Sites 982
and 925 spanning the late Miocene to early Pliocene “biogenic bloom” interval (Diester-Haass et al., in preparation).
Paleoproductivity is calculate using the relationship of Herguera (2000): paleoproductivity = 0.4 x water depth x
BFAR0.5. The gray horizontal bar in (B) outlines modern productivity in these regions (Antoine et al., 1996).
Preliminary low resolution Sr/Ca results using the entire coccolith size fraction (0-20μm)
from North Atlantic Site 982 illustrate that there is encouraging agreement between the bottom up
and top down perspective (Figure 8, next page). Site 982 coccolith Sr/Ca ratios increase from 9 to 5
Ma, the trend being paralleled by an overall increase the calculate paleoproductivity during this
interval of time. With a few exceptions, the two proxies agree on the finer scale as well. Maxima in
the paleoproductivity tend to be accompanied by maxima in the DSr (e.g., at 7.7 Ma, 6.9 Ma, 5.2 Ma,
and 5.0 Ma, Figure 8). In fact, the relationship between the two records is quite good with a
15
correlation coefficient of 0.46 (Figure 9); which is significantly different from zero at the 5% level.
Interestingly, the magnitude as well as the timing of the single-point maximum in the DSr at 4 Ma
recorded in the coccolith DSr agrees very well with the DSr maximum derived from fine fraction
sediment in the Indian Ocean (e.g., Baker et al., 1982; Billups et al., 2004) suggesting an extra
regional effect. These preliminary coccolith Sr/Ca data show much promise for interpreting the two
proxies in the context of primary paleoproductivity.
North Atlantic Site 982
40
0.26
top down (coccoliths)
bottom up (benthic forams)
35
0.25
0.24
25
0.23
DSr
Paleoproductivity
30
20
0.22
15
0.21
10
0.20
5
0.19
0
3
4
5
6
7
8
9
Age (Ma)
Figure 8. Comparison of Site 982 paleoproductivity derived from benthic foraminiferal accumulation rates with the
DSr of coccoliths derived from coccolith Sr/Ca ratios (to account for changes in seawater Sr/Ca ratios). The
agreement between the two proxies appears to be quite good both over the long term as well as in the peak-to-peak
comparison.
40
r=0.46
Paleoproductivity
30
20
10
0
0.19
0.20
0.21
0.22
0.23
0.24
0.25
0.26
DSr
Figure 9. Scatter plot of the Site 982 DSr versus the benthic foraminiferal derived paleoproductivity. The
correlation coefficient (r) is 0.46, which is significant at the 5% level (t-test: t=2.26; tcritical at α0.025 =2.093).
16
6. Proposed Work (Table 1)
A total of 184 intervals (plus 10% duplicates) will be processed for coccolith Sr/Ca
measurements from Sites 982 and 925 (Table 1). After cleaning, each sample will be microfiltered
using 8 μm polycarbonate filter to separate the size fraction. We chose this particular size fraction in
accord with Stoll’s work, which shows significant differences in the 3-5 μm versus the 8-12 μm size
fractions. Thus for each interval, two Sr/Ca measurements will be made. The coccolith Sr/Ca
cleaning method has been published by Stoll and Schrag, 2000; and Stoll and Ziveri (2002). Stoll
(pers. comm.) has kindly provided us with a very detailed step-by-step methodology that includes
separating the sample into discrete size fractions, and, as the outside member of graduate student
Amanda Waite’s Thesis committee, is offering her continued support with this project. An outline of
this method is summarized in Table 2.
For each interval, Mg/Ca ratios will be measured to monitor the extent to which surface water
temperature changes may have affected the Sr/Ca ratios. Mg/Ca cleaning will follow the “Mg
protocol” as outlined by Rosenthal et al. (2004), which includes several sonication steps to remove
fine fraction material, an oxidating step using hot buffered peroxide to remove organic matter as well
as an acid rinse. Dissolved samples will be sent for analysis using ICP-MS to the Oxford University
(see attached letter from R. Rickaby).
Smear slides will be prepared for each interval to monitor the down core changes in species
assemblages. To this end we will adopt the method of Backman and Shackleton (1982) and count the
number of coccoliths belonging to the dominant species per mm-2 on the separated coccolith fraction
on a smear slide. Graduate student A. Waite has already learned to prepare these slides and to
identify the species. S. Gibbs (a former graduate student of N. Shackleton) will help in statistically
quantifying the species counts (see attached letter of support from S. Gibbs).
17
I have all major equipment (e.g., a centrifuge, shaker, and a petrographic transmitting light
microscope with 1000x magnification). The only piece of equipment that needs to be purchased is a
laboratory counter for counting coccoliths to determine changes in species assemblages.
7. Significance
Coccolith Sr/Ca ratios are a relatively new proxy for paleoproductivity; benthic
foraminiferal accumulation rates a comparatively well-established one. Each approach is limited
by its own set of uncertainties. Together, however, the two proxies should yield a coherent
picture of paleoproductivity if our understanding of each is adequate. This study explores
whether a simple link can be made in the geologic record between surface ocean production of
organic matter and deep ocean consumption, which would strengthen reconstructions of
paleoproductivity. Preliminary results advocate such a simplistic link thereby lending support for
the use of these proxies despite the many uncertainties.
The results from this study are important to the paleoceanographic community because
they will advance our understanding of the coccolith Sr/Ca and BFAR proxies. If the preliminary
results are confirmed by the proposed study, combining the two proxies will yield a powerful
means to reconstruct paleoproductivity (coccolithophorid and other). This is important to the
paleoclimatologist because primary productivity determines the partitioning of CO2 between the
atmospheric and oceanic reservoirs. It is important to the objectives of the Petroleum Research
Fund because primary marine productivity is the source of potential hydrocarbon reservoirs.
18
Table 1. Summary of proposed work.
Site
temporal
Number of
*Number of
*Number of
resolution
intervals to be
Sr/Ca analyses
Mg/Ca
processed
(for two size
analyses
fractions)
982
50-80 kyr
96
211
105
925
50-80 myr
88
194
96
Total
184
405
201
*Number includes 10% duplicate values
19
Table 2. Processing bulk sediments for coccolith Sr/Ca analysis (Stoll, pers. comm.)
Step
1
process
-wet sieve ~ 1 g bulk sediment through a 20μm sieve using ethanol to
separate the foram and foram fragments from clay and coccolith fraction
-pour into 15ml centrifuge tube
-centrifuge or let settle overnight
2
-to the fine fraction, add 50% of 2% sodium hypochlorite and 30%
hydrogen peroxide to oxidize the organic matter
-react in a beaker for 1 hour, add 1 ml bleach and sonicate repeatedly.
-recover the sediment using a vacuum filter (0.45 or 1 μm cellulose nitrate)
3
-to the fine fraction, add ~4ml of “MNX” (50g hydroxylamine
hydrochloride in 400 ml NH4OH and 600ml DDH2O) to reduce Fe and Mn
oxyhydroxides
-shake for ~12 hours, centrifuge and siphon off the liquid
4
-to the fine fraction, add 4-10 ml “IONX” (65ml NH4OH in 1 L DDH2O) to
remove exchangeable Sr
-shake for ~2-24 hours, centrifuge and remove liquid
5
-rinse with DDH2O, shake for 2 hours, centrifuge and siphon off the liquid
6
-microfilter using a 8 μm polycarbonate filter and retain both fractions.
7
-dissolve in 2% HNO3, to ensure complete dissolution, centrifuge, siphon
off the liquid and transfer into clean microcentrifuge tubes for analyses
using ICP-AES.
20
References
Anand, P., H. Elderfield, and M. H. Conte, Calibration of Mg/Ca thermometry in planktonic
foraminifera from a sediment trap time series, Paleoceanography, 18, doi.:
10.1029/2002PA000846, 2003.
Andersson, C., and E. Jansen, A Miocene (8–12 Ma) intermediate water benthic stable isotope record
from the northeastern Atlantic, ODP Site 982, Paleoceanography, 18, 10.1029/2001PA000657,
2003.
Antoine, D., J.-M. Andre, and A. Morel, Oceanic productivity, 2, Estimation at global scale from
satellite (coastal zone color scanner) chlorophyll, Global Biogeochemical Cycles, 10, 57-70,
1996.
Backman, J., and N. J. Shackleton, Quantitative biochronology of Pliocene and early Pleistocene
calcareous nannofossils from the Atlantic, Indian and Pacific Oceans, Mar. Micropaleontol. 8,
141-170, 1983.
Balch, W. M., and K. Kilpatrick, Calcification rates n the equatorial Pacific along 140°W, Deep Sea
Research II, 43, 971-993, 1996.
Baker, P., J. M. Gieskes, and H. Elderfield, Diagenesis of carbonates in deep-sea sediments-evidence
from Sr/Ca ratios and interstitial dissolved Sr data, J. Sediment. Petrol., 52, 71-82, 1982.
Berger, W., and G. Wefer, Export production: seasonality and intermittency, and paleoceanogrpahic
reconsturction, Palaeogeogr. Palaeoclimatol. Palaeoecol. 89, 245-254, 1990.
Berger, W. H., R. M. Leckie, T. R. Janecek, R. Stax, and T. Takayama, Neogene carbonate
sedimentation on Ontong Java Plateau: Highlights and open questions, Proc. Ocean. Drill.
Program Sci. Results, 130, 711-744, 1993.
21
Billups, K., R. Rickaby, and D. P. Schrag, Cenozoic pelagic Sr/Ca records: Exploring a link to
paleoproductivity, Paleoceanography, 19, doi.:10.1029/2004PA001011, 2004.
Boyd, P. W., and P. P. Newton, Does planktonic community structure determine downward
particulate organic carbon flux in different oceanic provinces? Deep-Sea Research I, 46, 63-91,
1999.
Dekens, P. S., D. W. Lea, D. K. Pak, and H. J. Spero, Core top calibration of Mg/Ca in tropical
foraminifera: Refining paleotemperature estimation, Geochem. Geophys. Geosyst., 3, 1022,
doi.: 10.1029/2001GC000200, 2002.
Dickens, G. R., and R. M. Owen, The latest Miocene-early Pliocene biogenic bloom: a revised Indian
Ocean perspective, Mar. Geol., 161, 75-91, 1999.
Diester-Haass L, P.A. Meyers, L. Vidal, The late Miocene onset of high productivity in the Benguela
Current upwelling system as part of a global pattern, Marine Geology, 180, 87-103, 2002.
Diester-Haass, L., P. A. Meyers, and T. Bickert, Carbonate Crash and Biogenic Bloom in the
Late Miocene: Evidence from ODP Sites 1085, 1086 and 1087 in the Cape Basin,
southeast Atlantic Ocean, Paleoceanography,19, PA1007 10.1029/2003PA000933, 2004.
Diester-Haass L., K. Billups, K. Emeis, In search of a global biogenic bloom, submitted to
Paleoceanography.
Farrell, J.W., I. Raffi, T. R. Janecek, D. W. Murray, M. Levitan, K. A. Dadey, K.-C. Emeis, M. Lyle,
J.-A. Flores, S. Hovan, 1995. Late Neogene sedimentation patterns in the eastern equatorial
Pacific. In: Proc. ODP, Sci. Res., 138, College Station (TX), Ocean Drilling Program. 717-756.
Francois, R., S. Honjo, R. Krishfield, and S. Manganini, Factors controlling the flux of organic
carbon to the bathypelagic zone of the ocean, Global Biogeochemical Cycles, 16, 1087-1097,
2002.
Gibbs, S., N. J. Shackleton, and J. Young, Orbitally forced signals in mid-Pliocene nannofossil
assemblages, Marine Micropaleontol., 51, 39-56, 2004.
22
Grant, K. M., and G.R. Dickens, Coupled productivity and carbon isotope records in the southwest
Pacific Ocean during the late Miocene-early Pliocene biogenic bloom, Palaeogeogr.
Palaeoclimatol. Palaeoecol., 187, 61-82, 2002.
Hampt-Andreasen, G., and M. L. Delaney, Lithologic controls on calcite recrystallization in
Cenozoic deep-sea sediments, Marine Geology, 163, 109-124, 2000.
Herguera, J.C., and W. A. Berger, Paleoproductivity from benthic foraminifera abundance: glacial to
postglacial change in the west-equatorial Pacific, Geology, 19, 1173-1176, 1991.
Herguera, J.C., Last glacial paleoproductivity patterns in the eastern equatorial Pacific: benthic
foraminifera records, Mar. Micropaleontol., 40, 259-275, 2000.
Hodell, D. A., J. H. Curtis, F. J. Sierro, and M. E. Raymo, Correlation of late Miocene to early
Pliocene sequences between the Mediterranean and North Atlantic, Paleoceanography, 16,
155-163, 2001.
Lampit, R. S., and A. N. Antia, Particle flux in deep seas: regional characteristics and temporal
variability, Deep Sea Research I, 44, 1377-1403, 1997.
Lear, C. H., H. Elderfield, P.A. Wilson, A Cenozoic seawater Sr/Ca record from benthic
foraminiferal calcite and its application in determining global weathering fluxes, EPSL, 208,
69-84, 2003.
Levitus, S., and T. P. Boyer, World Ocean Atlas 1994, vol. 4, Temperature, NOAA Atlas NESDIS 4,
129 pp., Nat. Oceanic and Atmos. Admin., Silver Spring, Md., 1994.
Nees, S., Late Quaternary palaeoceanography of the Tasman Sea: the benthic foraminiferal view,
Palaeogeogr. Palaeoclimat. Palaeoecol., 131, 365-389, 1997.
Peterson, L. C., D. W. Murray, W. U. Ehrmann, and P. Hempel, Cenozoic carbonate accumulation
and compensation depth changes in the Indian Ocean, in Synthesis of Results from Scientific
Drilling in the Indian Ocean, Geophys. Monogr. Ser, 70, pp. 311-333, AGU, Washington, D.C.,
1992.
23
Rickaby, R. E. M., D. P. Schrag, I. Zondervan, and U. Riebesell, Growth-rate dependence of Sr
incorporation during calcification of Emiliani huxleyi, Global Biogeochemical Cycles,
10.1029/2001GB001408, 2002.
Rosenthal, Y., S. et al., Inter-laboratory comparison study of Mg/Ca and Sr/Ca, measurements in
planktonic foraminifera for paleoceanographic research, Geochemistry, Geophysics,
Geosystems, 5 (4), 10.1029/2003GC000650, 2004.
Schmiedl, G., and A. Mackensen, Late Quaternary paleoproductivity and deep water circulation
in the eastern South Atlantic Ocean: evidence from benthic foraminifera, Paleogeogr.
Paleoclimat. Paleoecol., 130, 43-80, 1997.
Shackleton, N.J., and S. Crowhurst, Sediment fluxes based on orbitally tuned time scale 5 Ma to 14
Ma, site 926, In Shackelton, N. J., Curry, W. B., Richter, C., and Bralower, TJ., (Eds.) Proc.
ODP, Sci.. Results:, 154, College Station, TX (Ocean Drilling Program), 1997.
Shipboard Scientific Party Leg 162, Site 982, Proc. ODP, Init. Repts., 162, College Station, TX
(Ocean Drilling Program), 91-138, 1997.
Shipboard Scientific Party Leg 154, In Curry, W. B., Shackleton, N. J., Richter, C., wet al. (Eds.)
Proc. ODP, Init. Repts., 154, College Station, TX (Ocean Drilling Program), 1995.
Stanley, S. M., and L. A. Hardie, Secular oscillations in the carbonate mineralogy or reef-building
and sediment producing organisms driven by tectonically forced shifts in seawater chemistry,
Palaeogeography, Palaeoclimatology, Palaeoecology, 144, 3-1, 1998.
Stoll, H.M., and S. Bains, Coccolith Sr/Ca records of productivity during the Paleocene-Eocene
thermal Maximum from the Weddell Sea, Paleoceanography, 8, 10.1029/2002PA000875,
2003.
Stoll, M. H., Y. Rosenthal, and P. Falkowski, Climate proxies from Sr/Ca of coccolith calcite:
Calibrations from continuous culture of Emiliani huxleyi, Geochim. Cosmochim Acta, 66, 927936, 2002a.
24
Stoll, H. M., C. M. Klaas, I. Probert, J. R. Encinar, J. I. G. Alonso, Calcification rate and temperature
effects on Sr partitioning in coccoliths of multiple species of coccolithophorids in culture,
Global and Planetary Change, 34, 153-171, 2002b.
Stoll, H. M., P. Ziveri, M. Geisen, I. Probert, and J. R. Young, Potential and limitations of Sr/Ca
ratios in coccolith carbonate: new perspectives from cultures and monospecific samples from
sediments, Phil. Trans. R. Soc. Lond. A., 360, 719-747, 2002c.
Stoll, H. M., and P. Ziveri, Separation of monospecific and restricted coccolith assemblages from
sediments using differential settling velocity, Mar. Micro., 46, 209-221, 2002.
Stoll, H. M. and D. P. Schrag, Sr/Ca variations in Cretaceous carbonates: relation to productivity and
sea level changes, Palaeogeography, Palaeoclimatology, Palaeoecology, 168, 311-336, 2001.
Stoll, H. M. and D. P. Schrag, Coccolith Sr/Ca as a new indicator of coccolithophorid calcification
and growth rate, Geochemistry, Geophysics, Geosystems, 1, 1999GC000015, 2000.
Stout, P. M., Interstitial water chemistry and diagenesis of biogenic sediments from the eastern
equatorial Pacific, Deep Sea Drilling Project Leg 85, Mayer, L., P. Theyer, et al., Init.
Reports, DSDP, 85, Washington DC, US Govt. Printing Office, 805-820, 1985.
van der Zwaan, G. J., I. A. P. Duijnstee, M. den Bulk, S. R. Ernst, N. T. Jannink, and T. J.
Kouwenhoven, Benthic foraminifers: proxies or problems? A review of paleoecological
concepts, Earth Sci. Rev., 46, 213-236, 1999.
Wilkinson, B. H., and T. J. Algeo, Sedimentary carbonate record of calcium-magnesium cycling,
American Journal of Science, 289, 1158-1194, 1989.
Yasuda, M., Berger, W.H., Wu, G., Burke, S., and Schmidt, H., Foraminifer preservation
record for the last million years: Site 805, Ontong Java Plateau, Proc. Ocean. Drill.
Program Sci. Results, 130, 491-508, 1993.
25